The country's two largest school districts have hit the brakes on generative artificial intelligence, blocking student access for the upcoming school year. Driven by mounting evidence of racial bias and a desire to preserve human connection, school leaders are reconsidering how fast these digital tools should enter classrooms. This pushback changes the conversation around education technology.
What Happened
Ahead of the 2026-2027 school year, New York City Public Schools enacted a one-year moratorium on student-facing generative AI tools for pre-kindergarten through eighth-grade students, according to the NYC Mayor's Office. The policy affects roughly 600,000 public school students and blocks approximately 40 educational programs, as reported by Reuters. While older high school students may still use AI in limited pilots, younger children will return to traditional, human-led instruction.
Shortly after, Los Angeles Unified School District implemented an even broader ban. According to TechPolicy.Press, the district blocked all of its 378,000 students from using generative AI on school-issued devices, including helper tools embedded within Google Classroom.
Meanwhile, state legislatures are writing new rules to regulate these tools. Oklahoma recently enacted Senate Bill 1734, known as the Responsible Technology in Schools Act. According to The Leveraged Years, the law establishes a mandatory review system where educators must approve any AI-generated content before it reaches students. Under the law, schools are also explicitly barred from using AI as the primary basis for grading, student discipline, or grade promotion, as reported by NewsOn6. School boards have before the 2027–2028 school year to formally adopt these protective policies.
The Bigger Picture
These decisions come as academic research reveals flaws in automated learning tools. A study published in Education Week found that AI writing coaches systematically alter their feedback depending on a student's race and gender. When demographic data is known, the software modifies its tone and recommendations, introducing bias into what is supposed to be objective feedback.
Perhaps most troubling, scholarly articles published by Oxford Academic warn that classroom AI models are creating 'synthetic achievement gaps' that disproportionately disadvantage Black students. Research in MDPI similarly shows that automated grading algorithms do not evaluate work the same way human teachers do, often repeating or worsening existing grading discrepancies.
As we previously reported, the vast majority of parents and voters are demanding strict regulations to prevent tech companies from testing unproven software on children. While some states have embraced open access, such as Missouri's effort to partner with Google to offer free certificates, experts advise caution. Writing for the Hechinger Report, a former federal education expert argued that school districts must evaluate AI for racial and cultural bias using the same rigorous screening they already apply to data security and student privacy.
What This Means for Families
For parents, these policies mean children will bring home schoolwork that relies less on automated software. While some districts believe this pause will keep students safe, it also changes the classroom workflow. In Oklahoma, teachers now face the added task of vetting every piece of AI content, which could increase their workload and shift their attention away from direct instruction. Parents should also prepare for more school-level debates on how school surveillance and technology outpace data privacy protections.
What You Can Do
- Ask your local school board if they run bias testing on digital curriculum tools before signing purchasing contracts.
- If your child uses writing assistants or digital tutors at home, review the software settings to disable sharing personal data or demographics.
- Talk to your children about how AI models make mistakes, and encourage them to view automated suggestions as a draft rather than a final grade.